If light bulbs only burn out when turned on, may I assume that the sudden jolt of electricity through the cold metal is what really wears out the bulb? Rather than say, just leaving the bulb on?
For example, I have a lightbulb hanging over my small dining room that is on a dimmer switch. I can’t remember replacing that bulb ever, despite using it everyday. The dimmer, i speculate, allows the electrcity to slowly ease into the bulb, as opposed to a qucik burst. So would it be cheaper to have most of my lights on a dimmer switch?
Ya’know I’ve never really thought about whether a dimmer would increase a lightbulb’s lifetime, but put like that probably yes.
Your correct about why light bulbs tend to go when they’ve only just been switiched on, the filament has much less resistivity when it’s cold which means more current will flow thorugh it and it’s much more likely to go. A dimmer is just a variable resistor which will increase the resistance in the circuit meaning less current will flow.
I believe that wear in a normal incandescent light bulb is a combination of two (or three) effects: thermal stress from the heating and cooling and sublimation of the filament. (Maybe the alternating magnetic field within the doubly wound helix might also create mechanical stress?)
Anyway, the first two factors ought to be smaller if you leave the light on a dimmer, as there will me less thermal fatigue (slower heating up) and lower temperature (A dimmer, even at maximum setting will not let the light shine as bright as if it was fully on.). When it comes to the magneto-mechanical stress, I’m less sure. I would imagine that the sharp switching noise of the triac would cause strong magnetic fields within the filament, which could then result in mechanical fatigue. This surely calls for further investigation!
Note though, that for halogen bulbs it’s not recommended to run them for an extended time on ‘half’. The halogen gas is only efficient at higher temperature, and if the gas doesn’t achieve this temperature the filament might sublimate away faster than if it was running at full brightness!
I think old dimmers may have been transformer based. I can’t imagine any dimmers ever being resistors – they’d get really, really hot. I think most dimmers are triacs or SCR pairs. They’re not good for inductive loads, so for things like ceiling fans transformer-based dimmers still exist.
Although having said that, it really makes me wonder how my X10 switch for inductive loads works? Seems lightweight to be a transformer, and there’s no electronic control of a transformer like there is for the triacs.
When I was on the stage crew in Jr. High School, late 1950’s, the dimmers for the stage lights were transformerbased as you said. They were autotransformers.
I’d think they now use triac based systems similar to those used in residences.
I don’t recall ever seeing transformer based dimmers for residental use. At least not on on the mass market,
Yes and no but mostly yes although somewhat no so I would say it is a qualified yes although i could argue it is a no because it is not 100% yes.
The problem is that the filament does not heat up evenly during the first fraction of a second so you get parts that heat up slower and do not limit the current in the part that heats up faster and that is where the filament fails. If you limit the initial current externally then you are preventing this effect from happening.
You are correct that the dimmers limits the initial current and protects the light bulb. Whether it makes sense to fit all lights with dimers depends on the cost of dimmers, the cost of lightbulbs and how often they fail. If you really don’t need the dimmer for any other reason and the lights are not failing too often then it is probably not worth the trouble.
Well, autotransformers have been around for decades, and (assuming it’s properly rated) a variable autotransformer (e.g. Variac[sup]TM[/sup]) can be used to vary the AC voltage to just about anything (including lamps). The advantage of an autotransformer is that it supplies a nice, clean, noise-free sine wave, something a phase-fired proportional controller can’t do. The disadvantages of an autotransformer are that it’s big and expensive. I believe fans use a multi-tapped autotransformer, but I’m not sure.